Embedding
What it means: An embedding is a numerical representation of a word, sentence, or document. It captures the meaning and context of the text in a form that an AI model can use for comparison. Analogy: Think of an embedding as a location on a map of ideas. Text with similar meanings sits close together. “Gas station” and “convenience store” aren’t the same words, but they land near each other. Why it matters: Embeddings make it possible to compare concepts based on meaning rather than just exact words. How it shows up in Gloo: When you upload content such as sermons, PDFs, curriculum, or articles to the Data Engine, Gloo turns it into embeddings for you, so Search can match on meaning. If you run your own search, the Embeddings API creates embeddings for any text you send. Go deeper: Embeddings: How AI Understands Meaning explains how embeddings capture meaning and power search.Vector
What it means: A vector is the list of numbers that makes up an embedding. It’s what AI systems store and compare. Example: A short sentence like “I feel happy” might turn into a vector like [0.12, -0.45, 0.88, …]. These numbers don’t mean much on their own, but together they let the system measure similarity with other sentences. How it shows up in Gloo: Content processed by the Data Engine is stored as vectors. Search compares a question’s vector with those vectors to find the most relevant passages. The Embeddings API returns vectors directly, so you can store and compare them yourself.Vector Store (or Vector Database)
What it means: A vector store is a special kind of database that stores embeddings (vectors) and can quickly find the ones that are most similar to a new query. Use case: Say you ask, “What’s our return policy?” The system turns your question into a vector, finds stored content with similar vectors, and retrieves those documents. A model can then write an answer based on them. Popular tools: Pinecone, Weaviate, FAISS, and Qdrant are examples of vector databases commonly used with AI systems. How it shows up in Gloo: The Data Engine keeps a vector store of each organization’s uploaded content, scoped to its Publishers. It lets Search and the grounded endpoints find the most relevant passages, scriptures, or policy documents for a question. Go deeper: Building a Knowledge Base walks through how content is indexed for search.Semantic Search
What it means: Semantic search finds results by meaning, comparing embeddings instead of looking for the exact words you typed. Example: If you search “how do I get a refund,” the system might return a document titled “return policy” even though the exact word “refund” isn’t used. How it shows up in Gloo: Gloo AI’s Search API is semantic, so people can ask questions in their own words and still find the right material. You can try it in the Studio Search Playground. Go deeper: The Retrieval Step explains how a question is matched to the most relevant passages.Retrieval-Augmented Generation (RAG)
What it means: RAG is a technique where the system first finds documents relevant to a question, then gives them to the model to write its answer. It combines search and generation. Why it matters: RAG lets a model use current or specialized information without retraining. Use case: A church could use RAG so an AI tool answers questions from its own materials, even though the model was never trained on them. For example, a site search could return passages from past sermons, or answer a volunteer’s question about a specific policy. How it shows up in Gloo: Gloo AI’s grounded endpoints, Grounded Completions and Grounded Responses, do RAG for you. They retrieve the most relevant passages from your uploaded content, give them to the model, and tell you whether the answer was grounded. Grounded Completions can also return citations to the sources it used. Go deeper: How RAG Actually Works walks through retrieve, augment and generate step by step.Knowledge Base
What it means: A knowledge base is the collection of documents or facts a system draws from. With embeddings and a vector store, an AI can search and use it. Why it matters: It’s the information people choose and maintain, and it gives AI responses reliable, specific context. How it shows up in Gloo: Content you upload to the Data Engine becomes your organization’s knowledge base on Gloo AI, organized by Publisher. You can list, edit, and delete items with the Content APIs, and search it or ground answers in it. Go deeper: Building a Knowledge Base covers choosing sources, chunking and metadata.Enrichment
What it means: Enrichment means adding metadata, structure, or tags to documents, before or after they’re embedded. Good metadata makes it easier to find the right content. Example: Before storing a sermon transcript, you might add the speaker’s name, date, Bible passage, and topic as enrichment fields. When users ask questions, the AI can prioritize the most relevant enriched results. How it shows up in Gloo: You can attach metadata, such as title, author, and date, to content you upload to the Data Engine, and update titles, tags, authors, and more later with the Edit Content API.What’s next
To try these ideas in code, build Semantic Search with the Embeddings API, and see Preparing Content for Ingestion before you upload your own content.What Are Agents and How Do They Make AI More Capable?
What are AI agents, and how do they plan, act, and work across tasks like humans do?

